Detent Escapement Wheel Replacement?

[mynikko]

Hi,

I am currently working on two detent chronometer pocket watches that suffer the same symptom: running too fast. One gains a few minutes in an hour while the other one basically running 2X the speed (i.e. gaining 30 mins in 30 mins).

I think the teeth on the escapement wheels are worn that pallet of the balance wheel would sometimes miss the teeth. When the pallet misses a teeth, it skips to the next teeth that it can touch.


(the wheel in the above picture is about 1cm in diameter, including the teeth)

Based on my observation, the one that gains only(?) a few minutes per hour has one teeth worn. That means the pallet only touches 14 out of the 15 teeth on the escapement wheel per cycle, so it is about 1/15 faster than normal. The other one has several teeth worn so the it basically runs at 2X speed because only about half of the teeth can touch the pallet.

My question is that: is there a way to fix the wheel? Or wheel replacement is the only option? I have searched a bit and I could not find any detent escapement wheel replacement for sell. I also browsed through the pinned thread "Chronometer Parts and Service Sources", and I am not sure if any of them provides custom made escapement wheel service.

Please enlight me if you have had similar experience.

 

[gmorse]

Hi mynikko,

Have you established whether it's the impulse pallet on the roller missing a tooth or the locking stone on the detent mis-locking?

If it's the latter, adjusting the banking so that the locking stone is slightly deeper in the wheel may help, but if it's the former a replacement wheel may be the only option. Have you asked anyone about making a new wheel? I know someone here in the UK who could make one, but I'm sure it would be costly and he has a long backlog of work.

Regards,

Graham

 

[DeweyC]

Graham,

I suspect at least one of the wheels can be saved. But like everyone else, I have work into next year.

 

[mynikko]

By removing the balance wheel, I can see the locking stone on the detent can stop every teeth on the wheel (at least when I manually advance the wheel tooth by tooth). The black stain on the escapement wheel is the maker I left to obverse whether a teeth is missed or not (in action). I probably will remove all the markers and try to identify which teeth are missing the roller. I am thinking once I have Identified the shorter teeth, maybe I can tweak them a little so the roller can touch them again. If not, it's probably lathe time for me.

 

[mynikko]

OK, I removed the detent and examined whether the every tooth on the escapement wheel can touch the roller stone. Strangely, all the teeth have sufficient length to touch the roller stone. I re-assembled the watch and looked closely again; I found that the locking stone would just keep two teeth per interval. That is the reason why the watch is running at 2X speed.

Given the fact that there are 15 teeth on the escapement wheel, if teeth length is the main issue, there is no way the watch can run at exactly twice the speed (you can either damage 7 or 8 teeth but it will never give you the exact 7.5 teeth needed to achieve the 2X effect). Now I am suspecting the issue is within the interaction between the detent and the roller. Maybe the spring of the detent is too weak that it does not bounce back fast enough so an extra tooth is skipped every cycle, but I doubt if a mere weak spring can perform such constant skipping though. My other guess is that the gold spring tip is too long but I really do not want to mess with the gold spring length.

 

[gmorse]

Hi mynikko,

From your latest picture, this appears to be a pivoted detent banking in a hole in the plate, (another clue is the curved horn), so there's no option to adjust the detent foot. However, if it is a pivoted type, have you looked at the state of its return spring? You certainly shouldn't try shortening anything until the cause has been firmly established.

Regards,

Graham

 

[mynikko]

Hello Graham,

My habit is that I do not perform irreversible work unless I am absolutely sure what I am doing. I really hate to rescue a watch that the previous worker introduced even more problems. Anyway, back to the topic.



My literature tells me that the gold spring tip (the D part) should not be touching the pivot and should be filed to the thinnest state (please correct me if I am wrong). I suspect the gold spring tip in my case is a bit too long. Also, my gold spring tip was not thinned and has sign of excess wear. I do not know if this is the exact cause, but this is the direction I am working on. I measured the thickness of the A part, it is about 0.02mm.



I do not know what the correct thickness should be (it also has to do with the material). What I have on hand is apiece of 0.025mm copper sheet. I will try to cut it and replace the existing gold spring to see if my theory is correct. At least I have the original gold spring to fall back to if everything failed.



Thanks & Best Regards,

Nikko

 

[gmorse]

Hi Nikko,

Can you confirm that the gold spring is in one piece with the return spring on the tail of the detent? It seems an uncommon way of doing it. If, as it seems the locking is consistently only happening on every other tooth of the escape wheel, then something is preventing the detent from moving the locking stone back in time to lock the following tooth. Is the tip of the gold spring touching anything other than the impulse jewel at any time? You could check this if it's hard to see by putting a smear of engineer's blue on the impulse roller then look for any streaks on the roller or traces of blue on the tip of the gold spring after it's run for a few seconds.

If it isn't touching anything it shouldn't then I think you need to look at the return spring part of it, which may have become distorted so that it isn't snapping the detent back fast enough to catch the following tooth.

Regards,

Graham

 

[mynikko]

Hello Graham,


(source: Escapement - Wikipedia )
By return spring, you meant part f of the above diagram right? In the above case, the gold spring and return spring are two separate components.

However, in my case, my gold spring and return spring are the same spring. It is the same design as the one shown here:

(source: Pocket chronometers with detent escapement)

Since they are in one piece, you can image the returning force is pretty weak as it only relies on that tiny 0.02mm copper sheet. That is why I choose a lightly thicker copper sheet (0.025mm) instead. I am not even sure if the current spring is the original one.


If my 0.025mm sheet is still too weak, then maybe I can use another thicker sheet for part A, a thinner sheet for part C, and fix them together using the screw at point B (will leave some extension to even the foce). Or just use a thicker one piece and file the C part to a thinner layer as the old textbook instructed me to.

I do not have engineer's blue on hand. I will get a red or blue marker tomorrow to do the testing.

Thanks & Best Regards,

Nikko

 

[gmorse]

Hi Nikko,

...By return spring, you meant part f of the above diagram right? In the above case, the gold spring and return spring are two separate components...

Yes indeed , because your diagram is of an Earnshaw type spring detent where the body of the detent is not pivoted but moves on that very thin spring which provides both returning force and accurate alignment. Almost all the pivoted detents I've seen have had a flat spiral hairspring to provide the returning force. However I agree that yours doesn't look as though it ever had that type of spring, but the present spring doesn't look like an original part, although it was clearly designed to work that way. I'm not sure that copper would be an adequate permanent solution but phosphor bronze might be better and more elastic.

Regards,

Graham

 

[Tom McIntyre]

Just thinking about the count logic of the escapement, the behavior indicates that a second tooth is passing the balance after it is given impulse from the prior tooth. That is to say, the return of the detent is too slow to capture the next escape wheel tooth. The pivoted detent depends on the return spring (whether spiral or flat) moving a fair bit of mass back into the path of the escape wheel after unlock and impulse.

I have not examined very many flat return spring pivoted detents, but my recollection is that they are steel to get the needed performance.

 

[gmorse]

Hi Tom,

I have not examined very many flat return spring pivoted detents, but my recollection is that they are steel to get the needed performance.

That sounds very plausible.

Regards,

Graham

 

[DeweyC]

I have avoided responding more to this thread because it is too easy to provide the un prepared w/ enough confidence that they create a real problem. If the teeth have been properly corrected, the issue will likely be that escapement itself is incorrectly set up. If the OP takes this breadcrumb and reads to the point of understanding, he may well be able to do this.

This is not about being knowledge protective. I only teach this under direct supervision where I can control the risks.

 

[mynikko]

So what you are saying is that the problem probably has nothing to do with the return spring but the escapement arrangement itself, but you are not telling because you only teach this in person.

It's OK. I can understand since you guys owe me nothing and are not obligated to answer any of the questions. Thanks for the hint, though.

 

[mynikko]

I probably know what you are talking about now (12), because such constant skipping usually means the arrangement was setup up (incorrectly) that way. However, knowing is one thing; execution is a whole other story.

 

[mynikko]

It has been half a year and I am reviving this old thread. I finally made this chronometer to work again. I actually knew what DeweyC was talking about, but I didn’t have the time & energy to work on this chronometer (was working on two repeaters and a Modernista pocket watch). I could just keep my mouth shut and pretend I didn’t start this thread, but I believe the purpose of forum is to share so I will share my findings here.


The key to the problem is the angle of the locking jewel. The locking jewel in the original setup was almost 45 degree while the ideal angle should be around 12 degree (that’s why I mentioned the number 12 in my previous response). When the angle is too wide, the locking jewel will always skip a tooth resulting the watch to run at double speed.


The fun part was to remove the locking jewel , re-insert it with a new angle and fix it with shellac. I had to repeat the process several times until the watch starts ticking at desired interval. It took me days to complete this task because I can only keep my focus at high level a few hours a day. I even designed and made new locking jewels so I can simplify this process. Because this chronometer uses a very thin spring instead of a pivoted coil spring, the weight of the locking jewel (plus shellac and wedge) also plays a factor here. I know very few people will have the need to remake the locking jewel, but if you do, my tip may help you.


I probably will avoid working on another chronometer form now on. Interestingly, I was chatting in my friend’s store last week, and we met a customer came in with a high grade detent chronometer that is beyond repair. That chronometer was actually ruined by one of the local watchmakers who clearly did not know what he was doing. What a waste. Anyway, thanks to everyone who had helped me here.

 

[DeweyC]

Mynikko,

Glad you were not offended and worked it out. Good job. Like you, I have seen a lot of work done by "excellent watchmakers" that almost ruined the piece. Last year I had a customer send me back his Nardin after he took it to to a local watchmaker who was a parishioner. He ruined a perfectly good Nardin detent which I had to replace with a modified Hamilton detent. Never told the owner; just said it was beyond his skill level.

The chronometer escapement is very simple (nothing more than a garden swing gate) but mistakes are exceedingly expensive.